Skip to main content

Full text of "Contingent reinforcement in budgeting and performance appraisals"

See other formats

Faculty Working Papers 


David J. Cherrington, University of Illinois 

J* Owen Cherrington, Pennsylvania State University 


College of Commerce and Business Administration 

University of Illinois at Urbana-Champaign 

College of Commerce and Business Administration 
University of Illinois at Urbana-Champaign 

April 19, 1973 


David J. Cher ring ton, University of Illinois 

J* Owen Cherrington, Pennsylvania State University 





David J. Cherrington 
Assistant Professor of 
Business Administration 
College of Commerce 


J. Owen Cherrington 

Assistant Professor of 


Pennsylvania State University 

Proof and manuscript copies should be sent to David J. Cherrington, 
Department of Business Administration, University of Illinois, 
Champaign, Illinois, 61820. 

Digitized by the Internet Archive 

in 2011 with funding from 

University of Illinois Urbana-Champaign 


Many important questions are frequently asked by organizational 
leaders regarding the use of quantitative measurements, eg. should for- 
mal performance appraisals be made on each employee, should minimum 
performance standards be established, should employ e'es participate in 
the preparation of budgets and performance standards, should a manage- 
ment by objectives program (MBO) be utilized to reward employees for 
achieving their objectives, etc.? These are important practical ques- 
tions to those who are faced with the day to day operations of an or- 
ganization. The research stimulated by these questions, however, seems 
to have raised more questions than answers. A manager who seeks a 
straightforward answer to whether he should have his workers participate 
in preparing the budget is often told that it depends. If he asks 
"What does it depend upon?" he is told "Just about everything." Research 
on these topics has been inconsistent and unclear. It is our contention 
that the proposed answers to these questions have been based upon re- 
search findings obtained in unique conditions of experimental control 
and the most salient considerations (viz.. the reinforcement schedules 
and the performance -reward contingencies) which should have been ab- 
stracted from these studies have been imbedded in the experimental 
Reinforcement Schedules 

The work of B. F. Skinner and the operant conditioners has been 
largely neglected in the areas of management, accounting, and organiza- 
tional behavior. Within the past few years, however, the principles 
of operant conditioning have begun to emerge from the learning theory 
literature and appear in other areas of industrial psychology and organ- 
izational behavior. 1 We are suggesting here that the principles of 


operant conditioning and reinforcement can be meaningfully applied to 
the processes of budgeting, performance standards, and performance 
appraisals to predict and control attitudes and behavior. 

The basic principles of operant conditioning center around the 
development of stimulus-response (S-R) bonds. One speaks of an S-R 
bond when a particular stimulus regularly elicits a particular res- 
ponse from an organism. In genera. le strength of an S-R bond is ■ 
increased when the occurrence of the response is followed by a reward 
and decreased when followed by punishment. In other words, the sti- 
mulus comes to elicit the response dependably because the response ob- 
tains for the organism something of positive value. 

To explain the reinforcing properties of rewards , moderate be- 
haviorists 2 have assumed the existence of primary drives such as hun- 
ger, thirst, sex, activity, curiosity, and removal of pain. Corres- 
ponding primary reinforcers are food, water, sexual activity, exercise, 
information, and freedom from pain. In addition to primary drives 
and reinforcers there are secondary drives and reinforcers which are 
learned by the organism because of the^r association with primary re- 
wards. Thus such diverse goals as money, status, promotions, verbal 
praise, associations with friendly co-workers, etc., can be concep- 
tualized as secondary rewards, linked by past associations with such 
primary rewards as food and water. 

Radical behaviorists 3 do not make a distinction between primary 
and secondary drives and rewards. Nor do they rely upon drives, needs, 
or other mentalistic processes to explain reinforcement. Instead, 
when a stimulus following an operant response increases the strength 
of the S-R bond, it is called a reinforcing stimulus. Thus a reinforcer 


is said to be positive when the effect of the operant response is 
experimentally observed to "produce" the reinforcing stimulus. In 
contrast, whenever the effect of an operant response is experimentally 
observed to "eliminate" a stimulus , that stimulus is called a negative 

The above description of operant conditioning is greatly simpli- 
fied and has not considered several salient aspects of conditioning, 
eg. the timing of the reward and the importance of immediate feedback 
on performance. A crucial aspect is the reward schedule , which is 
the frequency with which a reinforcer is administered. A fixed inter- 
val reward schedule is when the reinforcement is given after a speci- 
fied period of time. With this schedule the reinforcement is not con- 
tingent upon (ie. related to) a specific response but is only con- 
tingent upon the passing of time. A continuous reward schedule is 
defined by a 1:1 association between a response and the occurrence 
of a reward i.e., every time the response is emitted it is rewarded. 
With an intermittent reward schedule, only some of the specific res- 
ponses are followed by a reward. There are two basic ways in which 
intermittent reinforcement can be administered. Every nth response 
may be reinforced on a fixed ratio schedule, or an average of 1/n of 
the responses may be reinforced in a random pattern on a variable ratio 
schedule. Ratio schedules tend to generate a high rate of response; 
furthermore, the variable ratio schedule leads to a more durable res- 
ponse (i.e., does not extinguish as rapidly once reinforcement is ter- 
minated) than both the fixed- ratio and continuous patterns. Under a 
fixed interval schedule there is an uneven behavioral pattern that varies 
from a very slow, unenergetic response immediately following reinforce- 

ment to a fast, vigorous response immediately preceding reinforcement. 
Fixed interval schedules have also led to what has been called "super- 
stitious" behavior in animal studies. 1 * Numerous other reinforcement 
schedules have also been tested for their effectiveness in changing 
behavior, and organizational leaders responsible for constructing re- 
ward schedules ought to examine this information. 

Strategies for changing behavior can be implemented by punishment, 
extinction, and positive reinforcement. Punishment is a widely used 
technique for behavioral control and often has a rather immediate effect 
in stopping or preventing undesired responses. Punishment, however, 
has generally been considered an inefficient technique for controlling 
behavior for a number of reasons. Essentially, punishment causes dys- 
functional consequences, e.g., (a) it might inhibit a specific response 
but possibly for only a short time, (b) it does not produce the cor- 
rect response, (c) it might interfere with the acquisition of the. cor- 
rect response, or (d) it might lead to avoidance and dislike of the 
punishing agent. 

Extinction generates fewer dysfunctional consequences than punish- 
ment but still does not increase the probability that the correct res- 
ponse will be emitted. Positive reinforcement is most efficient in 
changing behavior for at least t; reasons. First, it increases the 
probable occurrence of the desired response. Second, the adverse emo- 
tional response* associated with punishment and extinction are apt to 
be reduced and favorable emotions may be developed. 

Since about 1958, the use of positive reinforcement has been exam- 
ined much more intensively to see how the principles of operant con- 
ditioning can be applied to social behavior outside the experimental 

• 5- 

laboratory. Most notably, a reinforcement analysis has indicated that 
the acquisition of numerous responses is consistent with the principles 
of operant conditioning, e.g» , perception, verbal conditioning, inter- 
personal attraction, attitude formation, conformity, and leadership. ^ 
Nord has attempted to apply the principles of operant conditioning to 
management and organizational behavior <as emphasized the value of 
variable reinforcement schedules in job design, reduced absenteeism, 
and training and compensation programs.* 
Appropriate Reinforcement 

The determination of what constitutes ' 'appropriate" reinforcement 
is necessarily a subjective judgement and numerous theories have been 
suggested which consider different factors which influence one's sub- 


jective evaluation. Cherrington, Reitz, and Scott have suggested 
that appropriate reinforcement is defined by the performance -reward 
contingency. Appropriately reinforced individuals were the rewarded 
high performers and the nonrewarded low performers. Inappropriately 
reinforced individuals were the nonrewarded high performers and the 
rewarded low performers . 

We have attempted to apply the concept of appropriate reinforce- 
ment to the budgeting process. Essentially, we postulate that when 
budgets are used s appropriate reinforcement is provided when an indi- 
vidual's reward is contingent upon some combination of estimated and/ 
or actual results, depending on his ability to control each. Rewards 
should be contingent on a budget only if the individual is free to set 
an estimate without undue pressure or interference from upper manage- 
ment. Inappropriate reinforcement can exist (a) when rewards are based 
upon the budget and the individual is coerced into submitting an unreal- 


istic budget, or (b) when he is required to set a budget but not re- 
warded for making it* Essentially, rewards for estimates and/or per- 
formance should be compatible with the level of emphasis placed upon 

In a laboratory study using 230 undergraduates, significatly 
higher measures of both performance ion were obtained 

under conditions which approximate-. nforcement. Three- 

person groups estimated ir performance and n performed a simple 
task over 6 periods. The Ss were under one of three reinforcement 
contingencies, and one of four conditions of budgetary control regarding 
the acceptance of the estimates.- 5 - As predicted, the highest perfor- 
mance and satisfaction scores were obtained by Ss in conditions where 

(a) points were given both for the number of items produced as well as 
the number estimated and the groups had complete freedom in setting 
their estimates, and (b) the groups were forced to accept a high esti- 
mate but there was not a sizable loss of points if the estimate was 
not reached. On the other hand, the lowest performance and satisfac- 
tion scores were obtained from (a) the groups who were allowed to sub- 
mit a very lenient (low) budget, regardless of point schedules and 

(b) from the groups who were required to submit an estimate and could 
submit any estimate they desired, but who were rewarded only for the 
number of items made and not for the number estimated. These results 
are illustrated in Table 1 which shows the deviations of the performance 
and satisfaction scores for each cell from the overall mean. ' 

Insert: Table 1 about here 


The reward schedule in which the highest performance and satis- 
faction scores were obtained provided the same number of total points 
for an item estimated and made as the other reward schedules. How- 
ever, the relative weights attached to the number of items estimated 
and made were different. These weights had been determined by a group 
of research assistants in a pilot group who were specifically asked 
to subjectively decide what they considered to be the most appropriate 
reward schedule based upon the nature of the task, what would be a 
reasonable incentive for increasing the group estimates, the risk 
associated with predicting their performance, the control they had 
over their output, etc. Thus, appropriate reinforcement was defined 
by the subjective decisions of this group regarding an adequate combina- 
tion of points for the amount estimated, the amount produced, and a 
penalty for overestimation. 
Budgets and Performance Estimates 

The question frequently asked is what are the effects of budgets 
on people. We do not think this question is the most meaningful one 
to ask for several reasons, most of v are associated with the lack 
of operational specificity regarding budgets. It is not budgets, per 
se, that have an effect on people, but the positive and negative rein- 
forcing consequences associated with budgets and their reward contin- 
gencies which need to be more clearly specified. 

The basic concept of budgeting involves estimating future perfor- 
mance , comparing actual results with the estimate and studying the 
differences (variances) between them. The term "budget" is not a uni- 
tary concept, however, but varies from organization to organization. 
Some of the factors which determine the type of budget and its effects 


include the type, of organization, the personalities of people affected 
by the budget , the leadership style of the organization leaders , the 
method of preparing the budget, and the desired results of the budgeting 

.Several aspects of budgets, both functional and dysfunctional, have 
been discussed in the literature. Some of the functional aspects of 
budgets are the positive contributions budgeting makes to planning, coor- 
dination, implimentation , and control, and it provides a basis by which 
efforts are rewarded .. More specifically, planning means establishing 
objectives in advance and identifying the steps by which the objectives 
are to be achieved. The planning process initiates coordination and 
clarification of subgoals to achieve major enterprise goals. A coor- 
dinated plan (or budget) provides a "blue print" for implimentation 
and control. As the plan is implimented, continuous feedback is needed 
to evaluate actual results. If actual results are compared with planned 
activities on a timely basis, corrective action may be taken as needed 
to achieve ultimate goals . Lowe and Shaw have described an organization 
in which the major planning decisions, resource allocations, and organ- 
izational control were centered around the budget. x 

Several dysfunctional aspects of budgets have also been discussed 
in the literature. The dysfunctional aspects observed by Argyris are 
{*.) budgets are used as a pressure device which tends to unite employees 
against management, (b) reward structures provide success to the finance 
staff by making factory personnel appear as failures, (c) emphasis on 
the departmental level to achieve the budget creates a department-cen- 
tered enterprise, and (d) supervisors use the budget as a means of jus- 
tifying and expressing their coercive leadership style and personality. 12 


Another dysfunctional aspect is the negative connotation associated 

with the term "budget" and the negative attitudes created by budgeting.^ 3 

Most of the dysfunctional consequences identified above appear 
to represent a failure of management to construct the appropriate re- 
ward schedules and reinforcement contingenci* It is not too surprising 
to learn that budgets are some times viewed by workers as a pressure 
device, because sometimes budgets are intentionally used to pressure 
employees for higher productivity. Rather than criticising them for 
uniting against management we perhaps ought to commend them for being 
so perceptive. While varying degrees of pressure can be exerted by 
budgets, this pressure is not necessarily good or bad. In the study 
reported earlier high budget pressures increased performance without 
reducing satisfaction, but only when the Ss were rewarded accordingly. ** 
The dysfunctional consequences of budget pressures seem to exist when 
budgets are used as punative devices for criticising performance and 
withholding rewards. The solution to this problem, as well as the 
problem of negative attitudes associated with budgets, is most likely 
to be found in the use of appropriate reinforcement for the desired 
responses. Beddingfield has suggested that the people affected by 
the budgets will respond with a proper attitude only if they are assured 
by experience that the use of the budget is to improve their operations 
and not to simply single them out as an inefficient element in the firm.^ 

le other dysfunctional consequences attributed to budgets are 
created by inadequate contingencies of reinforcement. When the success 
of the finance staff is at the expense of the production personnel, or 
when one department can maximize their rewards at the expense of another 
department, there quite clearly is a need to rearrange the reward con- 


tlngencies. If two departments are interdependent then their reward 
contingencies ought to be tied together to the extent that coopera- 
tion is required. On the other hand, if they are independent their 
reward contingencies should not be too closely tied together since 
such a situation leads to a loss of control and noncontingent rein- 
forcement . 
Budget Participation 

Much has been written relative to the effects of worker parti- 
cipation in the budgeting process. The primary purposes for solici- 
ting participation in the budgeting process are to (a) gain accep- 
tance of a plan of action, (b) improve morale among employees and 
toward management, and (c) increase productivity. 

Participation in the budgeting process has been found in some 
studies to increase the probability of its acceptance. Argyris , in 
a study of the effect of manufacturing budgets upon frontline super- 
visors, points out that "....goals are more often accepted if the 
individual members can come together in a group, freely discuss their 
opinions concerning these goals, and ta'.e part in defining the steps 
by which these goals will be accomplished." 16 Similar arguments have 

•n made by others, e.g., McGregor-!-^ In studying resistance to change 
by employees in a pajama manufacturing plant, Coch and French found 
that employees who participated or who were represented in a group 
meeting to discuss the need for change, new job plans, and piece rates 
had a more cooperative and permissive attitude toward making the change 
than those not represented. ° The belief that an individual will be 
more receptive and more favorably disposed to a procedure if it contains 
his ideas is common to several theories of management. ^ 


There is also evidence in the literature that participation 
improves morale. Coch and French found a much lower turnover rate, 
fewer grievances about piece rates, and less aggression against the 

supervisor as individual participation in planning job changes in- 

creased. French, Israel, and As in studying participation in a 

Norwegian factory found that worker participation increased job satis- 
faction by satisfying important needs such as the need to be valued 
and appreciated by others. 21 Active participation seems to be desired 
by employees. It makes them feel more a part of the activities, less 

dominated by a superior, more independent, and thus improves the 

employees attitude toward his job. 

Increased productivity may, but does not necessarily, result 

from worker participation. Prior research has been contradictory on 

this point. Coch and French found greater productivity as a result 

of worker participation. v However, in attempting to replicate the 

study in a different organization and culture, French, Israel, and As 


found productivity uneffected. The results of a field experiment 
conducted by French, Kay, and Meyer found no significant differences 
in goal attainment between participative ly and non-participatively 


set goals. *• Tli us , the literature suggests that there is no direct 
correlation between participation and improved productivity. 

It has been argued by some that participation exerts an impact 
on productivity chiefly through its effect on group cohesiveness. 
However, several studies have shown that there is not a direct rela- 
tionship between, group cohesiveness and productivity but that the 


relationship is mediated by the norms of the work group. If the 

group norms are counter-productive or inconsistent with the goals of 


the organization then increased group cohesiveness would most likely 
decrease productivity. 

Like the term "budgets," the term "participation" is subject to 
different empirical interpretations ana there are great differences 
in the situations in which it has t Led. One could attempt 
to develop a complete categorization of the various forms of partici- 
pation and a catalog of situations appropriate to each form. It 
seems more appropriate to us, however, to focus upon the positive or 
negative reinforcing properties of participation. 

Participation in the preparation of a budget is a form of job 
enrichment and, hence, is a potential positive organizational rein- 
forcer for most individuals. Katz and Kahn have argued that indivi- 
duals want to be involved in significant decisions which affect their 
own fate. 29 They attributed the failure of French, Israel, and As to 
find performance increases in a Norwegian shoe factory to the lack of 
freedom the workers were allowed to make significant decisions, and 
to the absence of accompanying rewards. Besides the major factor of 
influence in significant decisions, participation invokes powerful 
forces of self-expression and self-determination. Not only are people 
discussing important matters , but each individual is given a chance to 
express his own views and to persuade others. Hulin and Blood have 
suggested that job enrichment is a significant positive outcome for 
individuals who espouse the Protestant work ethic (which suggests that 
participation may not have acquired reinforcing properties for all 
individuals). For those individuals who value enriched jobs, parti- 
cipation in the budgeting process (as well as participation in other 
activities, e.g., decision making, and performance appraisals, etc.) 


should increase job satisfaction. Productivity, however, might in- 
crease, decrease, or stay the same depending upon (a) appropriate 
reinforcement for the actual level of performance, (b) the pressure 
(which is itself a form of positive or negative reinforcement) from 
either peer group norms or management for a high performance standard, 
and (c) the efficiency of the organizational structures or production 
processes adopted by the group as a result of their participation. 
Performance Appraisals 

The entire notion of appropriate reinforcement and performance- 
reward contingencies requires that performance be evaluated. Like 
the topics discussed above, however, performance appraisals produce 
both functional and dysfunctional consequences. 

At the beginning of this article the question was posed whether 
performance appraisals should be conducted. This is a moot question 
since they are, in fact, done. It isn't really a question of whether 
or not we should conduct performance appraisals, but how we do it 
more effectively. The demands to hire, fire, promote, and compensate 
all necessitate some form of evaluatioi . Supervisors have always 
evaluated their subordinates and it is no. doubt true that evaluations 
made in a random and unsystematic fashion, unrecorded and undefended, 
whether valid or not, have in the past been just as important in 
influencing personnel actions as evaluations made in a more formal 

Numerous systems have been suggested for evaluating performance. 3 ^ 
It is usually desirable to obtain objective data when possible and 
applicable. Objective measures include production data (e.g., number 
of items made, number of errors, volume of sales, commissions earned, 


number of publications, etc.) and personnel data (e.g., absences, tardi- 
ness, accident rates, number of suggestions, etc.). When objective 
measures do not exist or can't be obtained s subjective judgments can 
be obtained from peer ratings , suborc ratings or superior ratings . 
Such evaluations are subject to various weaknesses which can be quite 
serious in some conditions. aphic rating scales (where the rater 
makes his evaluations on a scale from 1 to n regarding various charac- 
teristics) have frequently been subject to (a) unreliability, (b) the 
halo effect, (c) leniency in judgments, and (d) errors toward the cen- 
tral tendency. Some of these weaknesses have been minimized or eliminated, 
however by employee comparison methods and forced-choice techniques as 
well as by more refined rating scales. Some of the most frequently used 
characteristics include quantity of performance, quality of perfor- 
mance, dependability, job knowledge, initiative, cooperation, and over- 
all performance. 

Perhaps the most important point to remember is that these evalua- 
tions are value judgments. When evaluating an employee's performance 
the rater is imposing his own value system of what is good or bad, 
valuable or worthless upon the judgments he makes. This suggests that 
serious thought be given to the characterisitics and dimensions upon 
which an individual is evaluated. Guion has suggested that all eval- 
uations be compared against the concept of an ultimate criterion (after 
Thorndike). 2 The ultimate criterion is not a measure; but it is an 
abstraction embodying everything that ultimately defines success on the 
job. As a theoretical guideline, Guion suggested that performance 
appraisals be free from the contamination of anything unrelated to the 
ultimate criterion but not leave anything related to it out. 


Several dysfunctional consequences associated with performance 
appraisals have been identified in ongoing organizations and described 
in the literature. 33 Labovitz has argued that "at the higher levels 
of organizational life a greater portion of an individual's perfor- 
mance becomes difficult to measure in quantitative terms. Technical 
demands placed on individuals at tech. can usually be 

measured by some objective standards, -while the executive or adminis- 
trative functions cannot. While executive or administrative func- 
tions are more difficult to quantify it is unnecessary despair to say 
that they can't be evaluated. When objective measures are not avail- 
able, subjective judgments can be obtained. When we say that super- 
visor A is better than supervisor B or that supervisor A is doing a 
good or poor job we are simply stating an attitude regarding their 
performance. Considering the amount of thought and research which 
has been devoted to attitude measurement , it is quite likely that 
these attitudes can be measured reliably. 

Furthermore , even though it is desirable to have quantitative 
measures , such as the number of units produced or errors made , there 
are other important dimensions in every job. Any evaluation or ap- 
praisal which only looks at the quantity and quality of performance 
in objective terms omits the spontaneous and innovative behaviors 
(discussed by Katz and Kahn) which are vital to organizational ef- 

The difficulties associated with measuring the diverse aspects of 
an administrator's job along with the dynamic nature of such a job 
have been cited as some of the major problems associated with management 
by objectives. 36 MBO provides for an individual to participate in 


selecting his own goals or objectives and assumes that he will work 
hard to accomplish them and be pushed internally by reason of his 
commitment. However, he is frequently evaluated on something less 
than the ultimate criteria, and his rewards are tied to the appraisal. 
Consequently, if behavior is a function of the reinforcement contingen- 
cies, as suggested earlier, only those activities included within the 
appraisal will be performed. Several activites vital to the effective, 
coordinated functioning of the organization are them omitted. The 
solution to this problem is to make the evaluation more closely appro- 
ximate the ultimate criterion. In addition, Levinson has suggested 
that every MBO and appraisal program should include regular appraisals 
of the manager by his subordinates and be reviewed by the manager's 
superior. If the supervisor is free to say that a subordinate was 
sometimes irresponsible and would not complete various odd jobs when 
asked, then the subordinate should be able to say when he evaluates 
his superior that sometimes his task instructions were so garbled that 
he frequently didn't know what to do. This points to the advantages 
of joint appraisal reviews when time permits. 

Thompson and Dalton have argued in favor of an objective-focused 

performance appraisal approach. Essentially what they advocate is 

specifying the specific objectives of each Individual in advance and 
holding them responsible for the outcomes during the performance ap- 
praisal,. This procedure avoids the negative consequences of a zero- 
sum situation in which some workers must be evaluated below average in 
order to have other workers evaluated above average. Since each man 
has different tasks and different objectives , there is no reason why 
he cannot experience some success. Marked improvement in the perfor- 


mance of one person does not automatically require that someone else 
must slip backward. 

Using performance appraisals to reward performance implies an 
underlying personal value to the effect that there should be a ba- 
lance between the contributions of the employee and his level of com- 
pensation. This personal value has underlied this entire discourse 
and it is appropriate that this fact be recognized. What we are basic- 
ally saying is that in return for the compensation an individual re- 
ceives from the organization it is appropriate to expect him to make 
a reliable contribution to the organization which will be evaluated and 
determine in part the magnitude of his reinforcement. Aside from 
this value is the idea that appropriate reinforcement which is contin- 
gent upon certain desired responses will increase satisfaction and 
performance as was discussed above. Someone with a different value 
system might feel differnetly about the need for a compensation-contri- 
butions balance, but should agree upon the effects of appropriate 
reinforcement . 



1. See, for example, Albert Bandura, Principles of Behavior Modi- 
fication (New York: Holt, Rinehart and Winston, Inc., 1969); 
Kenneth Goodall, "Shapers at Work," Psychology Today (November, 
1972), pp. 53-63; and Walter R. Nord, "Beyond the Teaching 
Machine: The Neglected Area of Operant Conditioning in the 
Theory and Practice of Management," Organizational Behavior and 
Human Performance , IV (1969), pp. 375-401. 

2. Defined by Sal vat ore R. Maddi , Personality Theories : A Compara- 
tive Analysis ; (Revised ed. , Homewood, Illinois: The Dorsey Press, 
1972), Ch. 11. 

3. Ibid. 

4. B.F. Skinner, Contingencies of Reinforcement: A Theoretical 
Analysis (New York: Applet on- Century- Crof ts , 1969). 

5. For a good collection of readings, see Elliot McGinnies and 
C.B. Ferster, The Reinforc e ment of Social Behavior (Boston: 
Houghton Mifflin, 1971). 

6. Nord, op, cit. 

7. J.S. Adams, "Inequity in Social Exchange," in L. Berkowitz (ed.), 
Advances in Experimental Social Psychology , Vol. II (New York: 
Academic Press, 1965), pp. 267-299; and Robert L. Opsahl and Marvin 
D. Dunnette, "The Role of Financial Compensation in Industrial 
Motivation," Psychology Bulletin , Vol. LXVI, No. 2 (1966), pp. 94-118. 

8. David J. Cherrington, H. Joseph Reitz, and William E. Scott, 
"Effects of Reward and Contingent Reinforcement on Satisfaction 

and Task Performance," Journal of Applied Psychology , Vol. LV, No. 6 
(1971), pp. 531-536. 

9. This study has been reported in greater detail in David J, 

Cherrington and J. Owen Cherrington, "Appropriate Reinforcement 
Contingencies in the Budgeting Process," (A paper presented 
at the Empirical Research in Accounting Conference, Chicago, 
May, 1973). 

10. The three reinforcement contingencies were "output-only" (5 
points per item), "budget-oriented" (1 point per item made, 4 
points per item estimated, and -6 points per item overestimated), 
and "output-budget" (4 points per item made, 1 point per item 
estimated, and -2 points per item overestimated). The four 
budgetary control conditions were "imposed" (only a high, stated 
estimate was acceptable^ "pseudo-participation" (a high but 
unstated estimate was required), "lenient" (a low, stated esti- 
mate was necessary \ and "group-based" (the group was free to 

set its own estimate).. 

11. A.E. Lowe and R.W. Shaw, "An Analysis of Managerial Biasing: 
Evidence from a Company's Budgeting Process," The Journ al of 
Management Studies , Vol V, No. 3 (October, 1968), pp. 304-315. 

12. Chris Argyris , "Human Problems with Budgets," Harvard Business 
Review ( January- February , 1953). pp. 97-110. 

13. Milton F. Usry, "Solving the Problems of Human Relations in 
Budgeting," Budgeting , Vol. XVI (November-December, 1968), 
pp. 4-6. 

14. Also reported in Cherrington and Cherrington, op. cit. 

15. Ronald Beddingf ield , "Human Behavior: The Key to Success in Bud- 
geting," Management Accounting (NAA ), Vol. LI (September, 1961) 
pp. 54-56. 


16. Argyris, op. cit., p. 108 

17. Douglas McGregor, The H uman Side of Enterprise (New York: 
McGraw Hill, 1961). 

18. Lester Coch and John R.P. French, "Overcoming Resistance to 
Change," Human Re lations , Vol. I (19*18), pp. 512-532. 

19. Rensis Lickert, The Human Organization (New York: McGraw Hill, 

20. Coch and French , op. cit. 

21. John R.P. French, Joachim Israel, and Dogfinn As, "An Experi- 
ment on Participation in a Norwegian Factory," Human Relations , 
Vol. XXII (1960) pp. 3-19. 

22. Reviewed and discussed by Victor H. Vroom, Some Personality Deter- 
minants of the Effects of Participation (Englewood Cliffs, N. J. ; 
Prentice-Hall, 1960). 

23. Coch and French, op. cit. 
2^. French, et. al. , op. cit. 

25. John R.P. French, E. Kay, and H.H. Meyer, A Study of Threat and 
Participation in a Performance Appraisal Situation (New York: 
General Electric Co., 1962). 

26. For example, two authors who have stated this argument are V. Bruce 
Irvine, "Budgeting: Functional Analysis and. Behavioral Implications," 
Cost and Management (Canada) , Vol. XLIV (March- April, 1970), pp. 6- 
16; and E.D. Smith, "Human Behavior: A Factor in Accounting," 
Management Services , Vol. II (September-October, 1965), pp. 53-58. 

27. Victor H. Vroom, "Industrial Social Psychology," in G. Lindzey and 
E. Aronson (eds.), The Handbook of Social Psychology , Vol. V (2nd 
ed, Reading, Mass.: Addison Wesley, 1969), pp. 196-268. 


28. G. Strauss, "Some Notes on Power Equalization," in H.J. 
Leavitt (ed.), The Social Science of Organizations (Engle- 
wood Cliffs, N.J. : Prentice -Hall, 1963), pp. 39-84. 

29. Daniel Katj: and Robert Kahn, The Social Psychology of Organ- 
izations (New York: John Wiley, 1966), Ch. 13. 

30. Charles Hulin and Milton Blood, "Job Enlargement, Individual 
Differences, and Worker Responses," Psychological Bulletin , 
Vol. LXIX, No. 1 (1968), pp. 41-55. 

31. Robert Guion, Personnel Testing (New York: McGraw Hill, 1965), 
Ch. 4. 

32. Ibid., p. 113. 

33. G.H. Labovitz, "More on Subjective Executive Appraisals: An 
Empirical Study," Academy of Management Journal , Vol. XV, No. 3 
(1972), pp. 289-304; Paul H. Thompson, Performance Appraisal : 
Some Unanticipated Consequences (Boston: Harvard Business School, 
Unpublished Doctoral Dissertation, 1969); Paul H. Thompson and 
Gene W. Dalton, "Performance Appraisal: Managers Beware," 
Harvard Business Review ( Jan uary- February , 1970), pp. 149-157. 

34. Labovitz, op. cit., p. 290. 

35. Katz and Kahn, op. cit. , Ch. 12. 

36. H. Levinson, "Management By Whose Objectives," Harvard Business 
Review (July-August, 1972), pp. 125-134. 

37. Thompson and Dalton, op. cit. 


Table 1 

Deviations from the overall means: performance 3 and 

satisfaction scores 


Group - 

Budget a 





ba , 





Perfor- ] 



Satis- Perfor- Satis- 
faction | mance faction 




■■ ■■ 1 


A + .58 

-. ... ... , 

A - .81 

A - .01 

A - .57 


+ 0.8 

B + .31 
C + .33 


B - .65 1 
C - .73 

- 0.2 

B + .05 
C - .17 

- 1.5 

B - .53 


C - .78' 


D + .48 

D - .84 

D - .07 

D - .53 



A + .03 

A - .12 

A - .11 



A - .11 


i + 2 - 2 

B .00 
C - .16 

+ 0.4 

B - .01 
C - .16 

+ 2.3 

B + .20 
C - .19 

- 2.2 


B - .33 
C + .14 


D - .07 

D - .09 

D + .03 


D - .09; 


A + .28 

-'■- • 

A + .30 

A + .47 

■•■ "•* "H 

A + .07 

Output £ 


I + 0.8 


B + .17 
C + .26 

+ 2.0 

B + .56 
C + .79 

+ 2.0 

B + .35 

C + .53 

- 1.9 


B - .11 


C + .12, 


D + .24 

D + .46 

D + .31 



D + .16 



a. The overall performance mean for all six periods was 21.9. 

b. Satisfaction scores were based on a 1 - 7 scale; 1 = unfavorable and 7 = favorable. 
The attitude measures are indentified by the following letters: 

A General Satisfaction (overall mean = 4.63) 

B Personal Competence (overall mean = 5.22) 

C Supervisor Consideration (overall mean = 5.02) 

D Supervisor Competence (overall mean = 4.61)